Nickel nanoparticles in hydrogen-transfer reductions: Characterisation and nature of the catalyst

Article abstract of DOI:10.1016/j.apcata.2010.01.044

Nickel nanoparticles, readily prepared by reduction of nickel(II) chloride with lithium and a catalytic amount of DTBB, have been used in the transfer hydrogenation of carbonyl compounds and have been fully characterised by different means. The reaction rate of the transfer hydrogenation was found to be dependent on the acetophenone and isopropanol concentration but independent on the amount of lithium chloride. The deactivation of the catalyst after reuse has been mainly attributed to surface oxidation but not to coke formation. All the experiments performed are in agreement with the process being of heterogeneous nature. The nickel nanoparticles unveiled a superior behaviour in comparison with commercially available nickel catalysts.

Full text of DOI:10.1016/j.apcata.2010.01.044

Applied Catalysis A: General 378 (2010) 42–51
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Nickel nanoparticles in hydrogen-transfer reductions: Characterisation and
nature of the catalyst
Francisco Alonso *, Paola Riente, Juan Alberto Sirvent, Miguel Yus *
Departamento de Qu ı´ mica Org a´ nica, Facultad de Ciencias, and Instituto de S ı´ ntesis Org a´ nica (ISO), Universidad de Alicante, Apdo. 99, 03080 Alicante, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
Nickel nanoparticles, readily prepared by reduction of nickel(II) chloride with lithium and a catalytic
amount of DTBB, have been used in the transfer hydrogenation of carbonyl compounds and have been
fully characterised by different means. The reaction rate of the transfer hydrogenation was found to be
dependent on the acetophenone and isopropanol concentration but independent on the amount of
lithium chloride. The deactivation of the catalyst after reuse has been mainly attributed to surface
oxidation but not to coke formation. All the experiments performed are in agreement with the process
being of heterogeneous nature. The nickel nanoparticles unveiled a superior behaviour in comparison
with commercially available nickel catalysts.
Received 2 October 2009
Received in revised form 21 January 2010
Accepted 29 January 2010
Available online 6 February 2010
Dedicated to Professor Pelayo Camps on
the occasion of his 65th birthday.
ß 2010 Elsevier B.V. All rights reserved.
Keywords:
Hydrogen transfer
Heterogeneous catalysis
Nickel nanoparticles
1
. Introduction
In the recent years, considerable attention has been devoted to
applied in the absence of any anti-agglomeration additive or
nucleation catalyst at room temperature. These nanoparticles
found application in different functional group transformations
[13] as well as in the hydrogen-transfer reduction of carbonyl
compounds and olefins [14], and reductive amination of aldehydes
[15]. We also discovered that nickel, in the form of nanoparticles,
the development of uniform nanometre-sized nickel nanoparticles
because of their unique properties and potential applications in a
variety of fields including electronics [1], magnetism [2], energy
technology [3], or biomedicine [4]. In comparison with the noble
metals, nickel nanoparticles have been much less studied in
catalysis, although they have found a particular application in the
growth of carbon nanotubes [5] as well as in a variety of organic
reactions [6]. The synthesis of nickel nanoparticles in the zero-
valence state is not trivial since they readily undergo oxidation,
consequently affecting their catalytic performance [7]. Nickel
nanoparticles are mostly synthesised by the chemical reduction of
a nickel(II) salt, with the polyol process [8] and hydrazine [9] or
sodium borohydride [10] reduction being the most practiced
methods. In general, the presence of an additive, as protective
agent, is necessary and a common feature in all these methodolo-
gies in order to prevent particle agglomeration.
can activate alcohols for the
a-alkylation of ketones and indirect
aza-Wittig reaction, with this being a potential alternative to
noble-metal-based methodologies [16]. These reactions involved
the hydrogen transfer from the alcohol to the intermediate
a,b-
unsaturated ketone or imine, respectively. Moreover, in contrast
with the use of noble-metal catalysts, the reactions proceeded in
the absence of any added ligand, hydrogen acceptor or base, under
mild conditions. Very recently, we have demonstrated for the first
time that nickel, in the form of nanoparticles, can promote the
Wittig-type reaction of primary alcohols and phosphorus ylides
[17]. Moreover, a series of polymethoxylated stilbenes as well as
resveratrol, DMU-212, and analogues, have been synthesised using
this novel Wittig-type olefination as the key step.
As part of our continuous interest on the preparation and
application of active metals [11], we reported the fast synthesis of
nickel(0) nanoparticles (NiNPs), from different nickel(II) chloride-
containing systems in THF, using lithium powder and a catalytic
amount of an arene as reducing agent [12]. The method was
It is noteworthy that the NiNPs were shown to be catalytically
superior to other forms of nickel in all the aforementioned
reactions. Of particular interest is their application to the catalytic
transfer hydrogenation with isopropanol of carbonyl compounds,
where the NiNPs could be reutilised several times maintaining a
high activity in a very simple reaction medium composed of NiNPs,
isopropanol and the substrate, in the absence of any base [14]. A
detailed characterisation of the NiNPs in this reaction medium,
however, has not been yet tackled. We wish to report herein a
*
Corresponding authors. Tel.: +34 965903548; fax: +34 965903549.
E-mail addresses: falonso@ua.es (F. Alonso), yus@ua.es (M. Yus).
0926-860X/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.01.044

F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
43
complete study on the characterisation of the NiNPs, utilised in
2.3.3. X-ray powder diffraction
transfer hydrogenation reactions with isopropanol as the hydrogen
donor, as well as a series of catalytic and kinetic experiments. Since
catalysis by metal nanoparticles can be considered a kind of ‘‘semi-
heterogeneous catalysis’’, at the frontier between homogeneous
and heterogeneous catalysis [18], the present study will also
include a series of tests in order to determine the catalytic nature of
the process.
The XRD diagrams were collected in the
Bruker D8 Advance X-ray diffractometer: Cu K
˚
= 1.5406 A; room temperature (25 8C); 2u = 4–80. The samples
u
–
u
mode using a
1
a irradiation,
l
were dried under vacuum and kept under an argon atmosphere.
2.3.4. Electron paramagnetic resonance
EPR spectra were recorded on a Bruker EMX EPR spectrometer
equipped with a Bruker ER041X microwave bridge X-VAN,
2
. Experimental
operating at X-band frequency (n = 9.8 GHz). The samples were
diluted with 0.5 mL of dry 2-propanol under an argon atmosphere
and the spectra recorded at room temperature.
2.1. General
Dry THF was directly used without any purification (Fluka,
2.3.5. Surface area
9
9.9%). Anhydrous nickel(II) chloride (Alfa Aesar, 98%), lithium
The surface area was measured by the BET (Brunauer–Emmett–
0
powder (MEDALCHEMY S. L.), 4,4 -di-tert-butylbiphenyl (DTBB,
Fluka), isopropanol (Panreac, 99.8%), acetophenone (Aldrich), and
triphenylphosphane (Aldrich) were commercially available of the
best grade and were used without further purification. The
chromatographic analyses (GLC) were performed with a Hewlett
Packard HP-6890 instrument equipped with a flame ionisation
detector and a 30 m HP-1 capillary column [0.32 mm diameter,
2
Teller) method using N at 77 K on a vacuum volumetric gas
sorption Autsorb-6 and Autosorb Degasser apparatus (Quanta-
chrome). The sample was carefully filtered through filter paper and
dried under vacuum prior to analysis.
2.3.6. Thermal programmed oxidation
TPO experiments were carried out in a U-shaped quartz reactor,
3
0
.25
nitrogen (2 mL/min) as carrier gas, Tinjector = 275 8C, Tcolumn = 60 8C
3 min) and 60–270 8C (15 8C/min).
m
m
film thickness (100% dimethylpolysiloxane)], using
using a 3% O
2
/He gas flow of 40 cm /min. The sample was
subjected to a thermal treatment at a constant heating rate (10 8C/
min), and oxygen consumption and CO evolved were monitored
(
2
as a function of temperature by on-line mass spectrometry
(OmniStar system, from Pfeiffer). The sample was previously
filtered, washed with isopropanol and ultra-pure water, dried
under vacuum for 24 h and kept under an argon atmosphere.
2
.2. Preparation of the catalyst
Nickel (II) chloride (130 mg, 1 mmol) was added over a
suspension of lithium (14 mg, 2 mmol) and DTBB (13 mg,
.05 mmol) in dry THF (2 mL) at room temperature under argon.
0
2.3.7. Raman spectroscopy
The reaction mixture, which was initially dark blue, rapidly
changed to black (ca. 5 min) indicating that the nickel(0)
nanoparticles were formed. Additional stirring of the resulting
suspension for at least 10 min is recommended before use.
Raman spectra were recorded on a 50 mW LabRam spectrome-
ter (Jobin-Ivon Horiba) equipped with a confocal microscope and
three laser excitation lines (l = 514, 632, and 785 nm). Detection
was done with a CCD detector (1064 ꢀ 256 pixels) refrigerated
with a Peltier system. The sample was filtered and dried under
vacuum prior to analysis.
2.3. Characterisation of the catalyst
i-PrOH (4 mL) was added to the aforementioned NiNPs
2.3.8. Thermogravimetric analysis
suspension and the mixture was warmed at 76 8C during 1 h
unless otherwise stated).
TGA was performed on a TG–DTA apparatus (Mettler Toledo
TGA/SDTA851e/LF/1600) connected to a cuadrupolar mass spec-
trometer (Pfeiffer Vacuum Thermostar GSD301T). The sample was
filtered and dried under vacuum prior to analysis.
(
2.3.1. Transmission electron microscopy
TEM images were recorded using a JEOL JEM2010 microscope,
equipped with a lanthanum hexaboride filament, operated at an
acceleration voltage of 200 kV. A drop of the nickel nanoparticle
suspension was added to a holey-carbon coated 300 mesh copper
grid allowing the solvent to evaporate before being introduced into
the microscope. X-EDS analyses were carried out with an Oxford
Inca Energy TEM100 attachment.
2.3.9. Inductively coupled plasma mass spectrometry
ICP-MS analyses were carried out on a Thermo Elemental VG
PQ-ExCell apparatus. The filtrate was evaporated and dissolved
with 6 mL of ultra-pure water.
2.4. Catalytic activity tests
2.3.2. X-ray photoelectron spectroscopy
2.4.1. Transfer hydrogenation of acetophenone
The XPS spectra were measured with a VG-Microtech Multilab
Unless otherwise stated, i-PrOH (4 mL) and acetophenone
(1 mmol) were consecutively added to a suspension of the as-
prepared NiNPs. The reaction mixture was warmed to the
indicated temperature and stirred for the specified time. The
resulting suspension was diluted with diethyl ether (20 mL),
filtered through a pad containing Celite, and the filtrate was dried
3
(
000 electron spectrometer using a non-monochromatised Mg-K
a
1253.6 eV) radiation source of 300 W and a hemispheric electron
analyser equipped with 9 channeltron electron multipliers. The
pressure inside the analysis chamber during the scans was about
ꢁ
7
ꢁ2
5
ꢀ 10 N m . After the survey spectra were obtained, higher
resolution survey scans were performed at pass energy of 50 eV.
The intensities of the different contributions were obtained by
means of the calculation of the integral of each peak, after having
eliminated the baseline with S form and adjusting the experimen-
tal curves to a combination of Lorentz (30%) and Gaussian (70%)
lines. All the bond energies were referred to the line of the C 1s to
4
over MgSO and analysed by GLC. For kinetic studies, the course of
the reaction was monitored by GLC analysis of aliquot samples
extracted at the specified time.
2.4.2. Poisoning tests
3
Hg(0) (5 or 10 mmol) or PPh (0.25, 0.5 or 1 mmol) were added
2
84.4 eV, obtaining values with a precision of ꢂ0.2 eV. The sample
over a suspension of the NiNPs (0.2 or 1 mmol) in the presence of i-
PrOH (4 mL), followed by the addition of acetophenone (1 mmol).
was handled under a helium atmosphere unless otherwise stated.

4
4
F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
Fig. 1. TEM micrographs of the NiNPs recorded at 1 h (a), 4 h (b), 8 h (c), and 24 h (d).
The reaction mixture was warmed to 76 8C, stirred at that
temperature for 24 h, and worked-up and monitored as said above.
and 8.3 keV (K lines) and 0.8 keV (L lines) (Fig. 3). The oxygen
detected could be attributed to partial oxidation of the nanopar-
ticles during the handling of the sample or to the presence of some
residual solvent.
3
. Results and discussion
3
3
.1. Catalyst characterisation
3.1.2. XPS
Fig. 4 shows the XPS spectrum of the NiNPs in the presence of
isopropanol under an inert argon atmosphere, prior to its
utilisation in a hydrogen-transfer reaction. The only peak at
852.2 eV corresponds to the Ni 2p3/2 level and it is characteristic of
Ni(0) [19]. However, the nanoparticles experienced oxidation upon
exposure to air as confirmed by the peak at 856.3 eV (Fig. 5). To
determine the unequivocal nature of the oxidised species is beyond
the aim of this study. Nonetheless, the peak at 856.3 could be
mainly ascribed to the Ni 2p3/2 level of Ni(II) in NiO, on the basis of
previous studies on the oxidation of nickel nanoparticles in air
[7,20].
.1.1. TEM and EDX
Droplets of the suspension containing the nickel nanoparticles
were analysed at different reaction times (Fig. 1). A typical TEM
micrograph and size distribution graphic after 1 h are depicted in
Figs. 1 and 2, respectively. Spherical, monodisperse, and highly
uniform nanoparticles were obtained with a narrow range of
particle size (0.75–2.88 nm, ca. 1.75 ꢂ 1.00 nm). Interestingly,
diameters <2 nm were measured for most of the nickel nanoparticles
(
ca. 75%). This result is quite different from that observed when the
nanoparticles were prepared at room temperature using THF as the
sole solvent (2.50 ꢂ 1.50, 25% NiNPs < 2 nm) but similar to that in the
presence of ethanol at room temperature after prolonged stirring
3.1.3. XRD
(
1.75 ꢂ 0.75, 78% NiNPs < 2 nm) [12b]. It is noteworthy that the
The XRD diffractogram of the NiNPs shows some broad and low
intensity peaks (Fig. 6). This behaviour could be attributed to the
sample being mainly amorphous and/or to the fact that the crystal
domains are <10 nm. Nevertheless, the peaks corresponding to
face-centred cubic nickel could be observed. A high intensity peak
morphology and size distribution was very similar irrespective of the
reaction time, as shown in Fig. 1, recorded at 1, 4, 8, and 24 h,
respectively. Therefore, the presence of isopropanol at 76 8C seems to
have a beneficial effect in the generation of the nickel nanoparticles as
regards their size, uniformity, and stabilisation.
at the angle 2
the reduction of NiCl
experiment of preparation of the NiNPs. The amount of lithium
u
2
= 15.35 could be assigned to NiCl . Very probably,
Energy-dispersive X-ray (EDX) analysis on various regions
confirmed the presence of nickel, with energy bands centred on 7.5
2
was not fully completed in this particular

F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
45
Fig. 2. Size distribution of the NiNPs determined by TEM at (a) 1 h, (b) 4 h, (c) 8 h, and (d) 24 h. The sizes were determined for 220, 187, 193, and 150 nanoparticles,
respectively, selected at random.
Fig. 3. EDX spectrum of the NiNPs.
metal added to the reaction medium is difficult to weigh accurately
14 mg of fluffy powder in a Schlenk). Consequently, small
variations on this weight can completely or partially reduce NiCl
The fact that NiCl was detected by XRD but not by XPS could be
due to the fact that NiCl is rather insoluble in THF/i-PrOH at room
diffraction pattern (SAED), which also suggest a very small size for
the crystalline domains and/or an amorphous character (Fig. 7).
(
2
.
2
3.1.4. EPR
2
EPR studies on nanosized nickel have been mainly conducted for
nickel oxide [22] and supported nickel [23], while its application to
unsupported nickel nanoparticles is little documented. We could
additionally confirm the presence of NiNPs in the zero-valence state
temperature and tends to settle down in the Schlenk. As a result,
droplets of the NiNPs suspension for XPS analysis would be exempt
of residual NiCl
2
, whereas the powder for XRD analysis would
contain this NiCl
2
.
by EPR. The spectra of NiCl
Al were alsorecorded asreferences(Fig. 8). Asexpected, theblank
experiment with NiCl gave no paramagnetic signal since in the
ground-state electronic configuration all the electrons are matched.
A similar magnetic response was observed for the nickel nanopar-
2 2
and commercially available Ni/SiO –
Although the sample preparation was carried out under an inert
atmosphere, it was exposed to air during the XRD analysis. The
presence of cubic NiO [21], however, was difficult to confirm due to
its minor proportion and/or amorphous character. The information
and conclusion derived from the XRD analysis are in agreement with
the presence of diffuse intensity rings in the selected area electron
2 3
O
2
2 2 3
ticles when compared with Ni/SiO –Al O , with the peak-to-peak
resonance line width ( pp) being 582 and 1193 G, respectively. The
DH

4
6
F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
Fig. 4. XPS spectrum of the freshly prepared NiNPs.
Fig. 7. SAED of the NiNPs.
Fig. 5. XPS spectrum of the NiNPs after exposure to air.
2 2 3 2
Fig. 8. EPR spectrum of NiNPs: (a) Ni/SiO –Al O , (b) NiNPs, (c) NiCl .
3.2. Catalytic and kinetic studies
All the studies were carried out using acetophenone as a model
substrate. In a preliminary study, the transfer hydrogenation was
analysed as a function of the reaction temperature (Fig. 9). Low
conversion was observed at 25 8C for short reaction times, which
increased up to 54% after 100 min. The conversion profiles at 40,
50, 60 and 76 8C show a progressive increment in the slope at short
reaction times, with the reaction at 76 8C reaching 88.5%
Fig. 6. XRD spectrum of the NiNPs.
effective g factor (geff) was experimentally determined as h
were
the resonance maximum occurs, h is the Planck’s constant, and
n
/
m
B
H,
n
is the microwave frequency, H is the magnetic field at which
m
B
is
the Bohr magneton. The geff values obtained were 2.22 for the NiNPs
and 2.15 for the Ni/SiO –Al catalyst.
2
2 3
O
3.1.5. Surface area
The surface area of a NiNPs sample was determined to be
2
3
3.04 m /g. This value is in agreement with those obtained for
other metal nanoparticles using the same methodology [24] but
greatly differs from that obtained for nickel nanoparticles
synthesised from the electric explosion of wire, the specific area
2
of which was 4.405 m /g [7].
Fig. 9. Transfer hydrogenation of acetophenone profiles at different temperatures.

F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
47
Table 1
Transfer hydrogenation of acetophenone with different nickel catalysts.
.
a
Entry
Catalyst
None
Product
Yield (%)
1
2
3
4
5
6
7
8
Acetophenone
Acetophenone
1-Phenylethanol
Ethylbenzene
Acetophenone
Acetophenone
Acetophenone
Acetophenone
100
100
94
b
NiNPs
NiNPs
c
Raney Ni
95
c
Ni-Al
100
100
100
100
d
2
Ni/TiO
c
3
Ni/SiO
2
–Al
2
O
c
NiO
a
GLC yield.
b
c
In the absence of isopropanol.
Commercially available catalysts.
d
Sample provided by the Inorganic Chemistry Department of the University of
Alicante.
conversion in only 8 min and a stationary conversion of ca. 89%
after 16 min. It is worthwhile mentioning that a similar maximum
conversion as that at 76 8C was recorded for the reactions at 40, 50,
and 60 8C after prolonged heating (90 min).
Fig. 10. Conversion and initial rate as a function of the amount of acetophenone.
We then compared the transfer hydrogenation of acetophenone
in the presence of different nickel catalysts at 76 8C (Table 1). Two
blank experiments, in the absence of isopropanol (only THF as
solvent) and in the absence of nickel gave the unchanged starting
material (Table 1, entries 1 and 2, respectively). It is noteworthy
that, under the conditions shown in Table 1, only the NiNPs were
able to reduce acetophenone to 1-phenylethanol (Table 1, entry 3),
whereas Raney Ni led to the hydrogenolysis product ethylbenzene
Table 2
Transfer hydrogenation of acetophenone at different NiNPs/acetophenone ratios.
[
25]. Some other nickel catalysts were completely inactive in this
reaction (Table 1, entries 5–8).
.
The concentration effects of acetophenone and isopropanol on
the initial rate were also examined. The conversion into 1-
phenylethanol as a function of time was plotted for a series of
reactions carried out with 1 mmol of NiNPs at a constant volume of
a
Entry
NiNPs/acetophenone (mmol)
t (h)
0.13
Yield (%)
1
2
3
4
5
6
7
1:1
89
87
7
1:5
1
1
1:10
1:10
1:20
0.5:1
0.25:1
2
mL of THF and 4 mL of i-PrOH with 0.25, 0.5 and 1 mmol of
30
24
2
36
0
acetophenone at 50 8C (Fig. 10). The initial rates were determined
for the individual runs and plotted against the amount of
acetophenone. Fig. 10 clearly depicts the reaction rate being
dependent upon the amount of acetophenone, albeit a negative
slope was obtained. This behaviour could be attributed to the
product 1-phenylethanol partially blocking the active sites by
adsorption on the nanoparticle surface. We also discovered that
high NiNPs/acetophenone molar ratios were crucial for the
reaction to occur (Table 2). Thus, 1:1, 1:5 and 0.5:1 NiNPs/
acetophenone molar ratios led to 1-phenylethanol in high yields
and short reaction times (Table 2, entries 1, 2, and 6, respectively),
whereas no reaction was observed when conducted with 1:20 or
86
0
24
a
GLC yield.
homogeneous ruthenium-pseudodipeptide catalysed asymmetric
transfer hydrogenation of ketones using i-PrONa/i-PrOH and also
observed that the initial rate was not directly proportional to the
hydrogen donor amount [26]. In both cases, a large excess of
isopropanol was utilised with respect to acetophenone.
In the same report as above, Adolfsson described an increase in
the degree of conversion after the addition of LiCl [26]. The authors
proposed that hydride transfer from isopropanol occurred in a
Meerwein–Ponndorf–Verley-type fashion [27] (i.e., direct hydro-
gen transfer through a six-membered transition state composed of
the hydrogen acceptor, metal, and hydrogen donor, without the
participation of metal hydride intermediates) but involving
ruthenium monohydride species. In our case, LiCl is in situ
generated during the formation of the nickel nanoparticles.
Previously performed deuteration experiments [14a], however,
indicated the participation of a dihydride-type mechanism, in
which the two hydrogens of the donor become equivalent after
being transferred to the metal to give the dihydride. These results,
together with the fact that the reaction rate was found to be
0
.25:1 molar ratios (Table 2, entries 5 and 7, respectively) and a
modest conversion was obtained with a 1:10 molar ratio after
prolonged heating (Table 2, entries 3 and 4).
We next analysed the effects of variation of the isopropanol
concentration. In this case, the transfer hydrogenation of
acetophenone was conducted with 1 mmol of NiNPs, 1 mmol of
acetophenone, 2 mL of THF and 1, 4, 6, and 8 mL of isopropanol at
5
0 8C (Fig. 11). The graphic showing the initial rate against the
amount of isopropanol unveiled an unexpected behaviour. Thus,
although the highest initial rate was recorded for the largest
amount of isopropanol (8 mL), both lower conversions and initial
rate were obtained with 4 mL in comparison with those obtained
with 1 mL of isopropanol. Recently, Adolfsson et al. studied the

4
8
F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
2 2
Fig. 12. TPO profile of deactivated NiNPs: (a) CO , (b) H O, (c) CO.
on the feedstock molecules, which react to form coke. The
deactivation of some common heterogeneous hydrogenation
catalysts (e.g., Pd/C and Raney Ni) after aging them in alcoholic
solvents has been recently reported [30]. Alcohols such as
methanol, ethanol, 1-propanol, and 2-propanol got decomposed
into carbon monoxide and carbonaceous species with the
consequent catalyst surface poisoning. Sinterisation, agglomera-
tion or size increase of the nanoparticles can also affect their
catalytic activity and, therefore, must also be analysed.
Fig. 11. Conversion and initial rate as a function of the amount of isopropanol.
In order to know whether the deactivation of the NiNPs in
hydrogen-transfer reactions was due to the formation of coke, a
deactivated sample, after reutilisation in the transfer hydrogena-
tion of acetophenone with isopropanol, was analysed by tempera-
ture programmed oxidation (TPO), Raman spectroscopy and
thermogravimmetry (TG).
TPO experiments have been found to be an efficient tool to
evaluate the presence of carbonaceous materials [31], either
derived from the support [32] or from a chemical reaction [33]. The
TPO profiles of the deactivated NiNPs in Fig. 12 show that CO
formation starts at 200 8C with a maximum production at 270–
20 8C. These data rule out the possible CO emission from coke
combustion, which typically starts above 400 8C [34]. A similar
2
3
2
profile to that of CO
2
is observed for H
2
O with a rate of formation
ꢁ1 ꢁ1
g . Therefore, organic compounds and
Scheme 1. (a) The dihydride-type mechanism for the transition-metal catalysed
transfer hydrogenation with isopropanol. (b) Deuterium labelling experiment in the
transfer hydrogenation of acetophenone.
<1.3 ꢀ 10ꢁ mmol s
4
molecular hydrogen, derived from the transfer hydrogenation
reaction or sample preparation and adsorbed on the catalyst
surface, could account for the results observed.
Raman spectroscopy has been widely applied to the study of
used catalysts [35]. It is known that the precise absorptions of coke
strongly depend on the compounds present in the reaction
medium and deactivation conditions. A typical Raman spectrum
of graphite or coke exhibits two bands: the first one is more intense
and located at about 1590 cm , and a second one, normally less
intense and broader centred at about 1355 cm
NiNPs showed a rather flat Raman spectrum (Fig. 13) either before
or after their deactivation. The bands at a lower frequency are
present in both the fresh and deactivated samples, and correspond
independent of the LiCl concentration, allows concluding that LiCl
is a spectator in our transfer hydrogenation reaction (Scheme 1).
3.3. Catalyst deactivation
One main advantage of the NiNPs in the transfer hydrogenation
ꢁ1
of carbonyl compounds was their reuse capability. Thus, in the
stoichiometric version (1:1 NiNPs/acetophenone ratio), the NiNPs
could be reutilised over four consecutive cycles without any
apparent loss of activity (94% average yield), the latter decreasing
up to the seventh cycle (59% yield) [14a]. In the sub-stoichiometric
version (1:5 NiNPs/acetophenone ratio, 20 mol% Ni), the NiNPs
could be reutilised three times without any apparent loss of
activity (88% average yield) with a final fifth cycle (77% yield)
ꢁ
1
[35,36]. The
ꢁ1
ꢁ1
to NiCl
2
(ca. 175 and 270 cm ) and NiO (ca. 500 cm ). The
could be attributed to an
presence of small amounts of NiCl
2
incomplete reduction during the formation of the NiNPs, whereas
partial oxidation of the NiNPs during the sample preparation,
involving filtration and washing in air, could account for the
formation of NiO.
Thermogravimetric analysis of a deactivated sample of the
NiNPs did not show any significant difference with respect to that
of freshly prepared NiNPs in the temperature range applied (room
temperature to 700 8C).
[
14b]. We were intrigued by this gradual loss of activity but, even
more, by the total deactivation after the above-mentioned cycles.
The physical deposition of matter and/or strong chemisorption of
species can induce important surface modification leading to
deactivation [28]. In particular, coke formation can occur by
catalytic decomposition of organic compounds even at low
temperatures [29]. The chemical nature of this deposit depends

F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
49
Table 3
Catalyst poisoning experiments in the transfer hydrogenation of acetophenone.
.
a
Entry
Acetophenone/
NiNPs (equiv.)
Poison (equiv.)
Product
Yield (%)
1
2
3
4
5
6
7
8
9
5:1
1:1
5:1
1:1
5:1
5:1
5:1
1:1
1:1
1:1
Hg (5)
Hg (5)
Hg (10)
Hg (10)
1-Phenylethanol
1-Phenylethanol
Acetophenone
Acetophenone
Acetophenone
Acetophenone
Acetophenone
Acetophenone
Acetophenone
Acetophenone
10
10
100
100
100
100
100
100
100
100
PPh
PPh
PPh
PPh
PPh
PPh
3
3
3
3
3
3
(1)
(0.5)
(0.25)
(1)
Fig. 13. Raman spectrum before and after deactivation of the NiNPs.
(0.5)
(0.25)
10
a
GLC yield.
3.4.1. Catalyst poisoning experiments
The mercury test has been largely exploited to identify
heterogeneous catalysts due to its ability to poison metal(0)
heterogeneous catalysts by formation of amalgam or adsorption on
the metal surface [38]. In fact, mercury(0) is probably the most
effective in poisoning metals that form an amalgam, such as Pt, Pd,
and Ni [39]. When acetophenone was subjected to the transfer
hydrogenation, in the presence of 20 or 100 mol% NiNPs and 5
equiv. Hg(0), 10% of the reduced product 1-phenylethanol was
obtained (Table 3, entries 1 and 2, respectively). By using double
amount of Hg(0), however, the reaction was completely inhibited
(Table 3, entries 3 and 4). Since mercury has been reported to
Fig. 14. XPS spectrum of deactivated NiNPs after reuse.
occasionally induce side reactions, the results of the tests are not
always conclusive enough to know the catalyst nature. Therefore, it
is convenient to confirm the poisoning results with a second test.
From the above experiments, we can conclude that the
deactivation of the catalyst after reuse is not due to coke formation
during the reaction, but more plausibly, to partial oxidation to
nickel(II) oxide. In fact, different deactivated samples examined by
XPS after reuse showed only one peak at 856.3 eV, corresponding
to the Ni 2p3/2 level of Ni(II) (Fig. 14). Although all the reactions
were performed under an inert atmosphere, isopropanol was not
exempt of oxygen when used in each cycle. Therefore, a certain
degree of surface oxidation could definitely stop the catalytic
activity of the NiNPs.
PPh
3
can also be used as a poison due to its strong affinity to metal
centres. If a catalyst is poisoned completely with <1 equiv. of PPh
that evidences the presence of a heterogeneous catalyst. Six
different experiments were carried out by varying the acetophe-
3
,
3
none/NiNPs ratio and the amount of PPh . We observed that
irrespective the acetophenone/NiNPs ratio, the transfer hydro-
genation was inhibited either by stoichiometric or sub-stoichio-
3
metric amounts of PPh (Table 3, entries 5–10).
3.4.2. Filtration test [37,38]
This test relies on a comparison of the catalytic activity before
and after filtering the active catalyst solution. For this purpose, a
3.4. Nature of the catalyst
2
standard NiNPs black suspension, prepared from NiCl , Li, and
Catalysis by metal nanoparticles can be considered a kind of
‘semi-heterogeneous catalysis’’, at the frontier between homoge-
DTBB (cat.) in THF/i-PrOH at 76 8C, was used. When this suspension
was filtered though a pad containing Celite, the resulting filtrate
was found to be inactive in the transfer hydrogenation of
acetophenone. In general, membrane filters with pores small
enough to exclude nanoparticles are slow and difficult to use.
Therefore, we attempted filtration of the above suspension via
‘
neous and heterogeneous catalysis [18]. Therefore, it is important
to determine whether the NiNPs are the true catalysts or a
reservoir for metal atoms that leach into solution [37]. In this
section, we will describe a series of control experiments in order to
know the true nature of the catalytically active species [38].
TEM analysis is a tool for the direct observation of NPs as
heterogeneous species present in the reaction medium. The
formation of NiNPs could be confirmed by TEM for every
hydrogen-transfer reaction practiced.
In the previously described catalytic studies the NiNPs
manifested some activity since the very beginning and, as
expected, no induction period was observed in the kinetic profiles,
this behaviour being more typical for NPs generated from a
homogeneous catalyst.
syringe through a commercially available 0.45
mm pore size filter
under argon. The resulting colourless filtrate was analysed by ICP-
MS showing the presence of 64 ppb Ni. This filtrate was shown to
be inactive when subjected to the transfer hydrogenation of
acetophenone (Scheme 2). The low content of nickel in the filtrate
could be due to nanoparticle agglomeration during the transfer of
the NiNPs suspension with the syringe. In fact, any reaction carried
out in a flask different to that in which the NiNPs were generated,
and to which part of the original suspension was transferred via a
syringe, failed.

5
0
F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
[
3] R. Karmhag, T. Tesfamichael, E. W a¨ ckelg a˚ rd, G. Nikalsson, M. Nygren, Sol. Energy
8 (2000) 329–333.
4] (a) I.S. Lee, N. Lee, S.R. Paik, T. Hyeon, J. Am. Chem. Soc. 46 (2007) 4630–4660;
b) S. Rodr ı´ guez-Llamazares, J. Merch a´ n, I. Olmedo, H.P. Marambio, J.P. Mu n˜ oz, P.
6
[
(
Jara, J.C. Sturm, B. Chornik, O. Pe n˜ a, N. Yutronic, M.J. Kogan, J. Nanosci. Nano-
technol. 8 (2008) 3820–3827.
[
[
5] J. Geng, B.F.G. Johnson, in: B. Zhou, S. Hermans, G.A. Somorjai (Eds.), Nanotech-
nology in Catalysis, vol. 1, Springer, Berlin, 2004, pp. 159–182.
6] (a) A. Houdayer, R. Schneider, D. Billaud, J. Ghanbaja, J. Lambert, Synth. Met. 151
(
(
2005) 165–174;
b) J. Park, E. Kang, S.U. Son, H.M. Park, M.K. Lee, J. Kim, K.W. Kim, H.-J. Noh, J.-H.
Scheme 2. Filtration test in the transfer hydrogenation of acetophenone.
Park, C.J. Bae, J.-G. Park, T. Hyeon, Adv. Mater. 17 (2005) 429–434;
(
(
(
(
c) M. Kidwai, N.K. Mishra, V. Bansal, A. Kumar, S. Mozumdar, Catal. Commun. 9
2008) 612–617;
d) A. Saxena, A. Kumar, S. Mozumdar, Appl. Catal. A: Gen. 317 (2007) 210–215;
e) A. Dhakshinamoorthy, K. Pitchumani, Tetrahedron Lett. 49 (2008) 1818–
1
823;
f) V. Polshettiwar, B. Baruwati, R.S. Varma, Green Chem. 11 (2009) 127–131.
7] P. Song, D. Wen, Z.X. Guo, T. Korakianitis, Phys. Chem. Chem. Phys. 10 (2008)
057–5065.
8] (a) L.K. Kurihara, G.M. Chow, P.E. Schoen, Nanostruct. Mater. 5 (1995) 607–613;
b) Y. Wada, H. Kuramoto, T. Sakata, H. Mori, T. Sumida, T. Kitamura, S. Yanagida,
Chem. Lett. (1999) 607–608;
(
[
[
5
(
(
(
c) M. Tsuji, M. Hashimoto, T. Tsuji, Chem. Lett. (2002) 1232–1233;
d) L. Bai, J. Fan, Y. Cao, F. Yuan, A. Zuo, Q. Tang, J. Cryst. Growth 311 (2009) 2474–
2479.
[9] (a) D.-H. Chen, S.-H. Wu, Chem. Mater. 12 (2000) 1354–1360;
(
(
(
b) S.-H. Wu, D.-H. Chen, J. Colloid Interface Sci. 259 (2003) 282–286;
c) S.-H. Wu, D.-H. Chen, Chem. Lett. (2004) 406–407;
d) K.H. Kim, Y.B. Lee, S.G. Lee, H.C. Park, S.S. Park, Mater. Sci. Eng., A 381 (2004)
337–342.
[10] (a) G.G. Couto, J.J. Klein, W.H. Schreiner, D.H. Mosca, A.J.A. de Oliveira, A.J.G.
Zarbin, J. Colloid Interface Sci. 311 (2007) 461–468;
Scheme 3. Summary of the tests to determine the nature of the catalysis.
(
b) P.K. Khanna, P.V. More, J.P. Jawalkar, B.G. Bharate, Mater. Lett. 63 (2009)
384–1386.
11] (a) F. Alonso, G. Radivoy, M. Yus, Russ. Chem. Bull., Int. Ed. 52 (2003) 2576;
b) F. Alonso, M. Yus, Chem. Soc. Rev. 33 (2004) 284–293;
(c) F. Alonso, M. Yus, Pure Appl. Chem. 80 (2008) 1005–1012.
12] (a) F. Alonso, J.J. Calvino, I. Osante, M. Yus, Chem. Lett. 34 (2005) 1262–
1
From the essays above it can be inferred that the transfer
hydrogenation of carbonyl compounds catalysed by NiNPs
occur under heterogeneous catalysis on the surface of the NiNPs
[
[
(
(
Scheme 3).
1263;
(b) F. Alonso, J.J. Calvino, I. Osante, M. Yus, J. Exp. Nanosci. 1 (2006) 419–
4
. Conclusion
433.
[
[
13] (a) F. Alonso, I. Osante, M. Yus, Adv. Synth. Catal. 348 (2006) 305–308;
(
(
b) F. Alonso, I. Osante, M. Yus, Synlett (2006) 3017–3020;
c) F. Alonso, I. Osante, M. Yus, Tetrahedron 63 (2007) 93–102;
Nickel nanoparticles, generated by reduction of nickel(II)
chloride with lithium in the presence of a catalytic amount of DTBB
and utilised in the transfer hydrogenation of carbonyl compounds
with isopropanol, have been characterised by TEM, EDX, XPS, XRD,
EPRand BET area. Thefastest conversion was reached at 76 8C (88.5%
in 8 min). The NiNPs havebeen found to be superior to commercially
available catalysts, including Raney nickel. The reaction rate is
dependent upon the amount of acetophenone with a negative slope,
but it is not proportional to the amount of isopropanol, and
independent on the LiCl concentration. Based on TPO experiments,
Raman spectroscopy, TG and XPS analysis, the deactivation of the
catalyst with reiterative reuse has been ascribed to surface
oxidation, while the formation of coke has been ruled out. Moreover,
the NiNPs have been demonstrated to be the true catalyst in this
reaction, the heterogeneous nature of the process being unequivo-
cally established on the basis of TEM, kinetic, poisoning, and
filtration experiments.
(d) F. Alonso, P. Riente, M. Yus, ARKIVOC iv (2008) 8–15.
14] (a) F. Alonso, P. Riente, M. Yus, Tetrahedron 64 (2008) 1847–1852;
(
(
b) F. Alonso, P. Riente, M. Yus, Tetrahedron Lett. 49 (2008) 1939–1942;
c) F. Alonso, P. Riente, M. Yus, Tetrahedron 65 (2009) 10637–10643.
[15] F. Alonso, P. Riente, M. Yus, Synlett (2008) 1289–1292.
[
16] (a) F. Alonso, P. Riente, M. Yus, Synlett (2007) 1877–1880;
b) F. Alonso, P. Riente, M. Yus, Eur. J. Org. Chem. (2008) 4908–4914.
(
[
17] (a) F. Alonso, P. Riente, M. Yus, Synlett (2009) 1579–1582;
(b) F. Alonso, P. Riente, M. Yus, Tetrahedron Lett. 50 (2009) 3070–3073;
(c) F. Alonso, P. Riente, M. Yus, Eur. J. Org. Chem (2009) 6034–6042.
[
[
18] D. Astruc, F. Lu, J.R. Aranzaes, Angew. Chem. Int. Ed. 44 (2005) 7852–7872.
19] R.P. Furstenau, G. McDougall, M.A. Langell, Surf. Sci. 150 (1985) 55–79.
[20] S. Lee, N. Lee, J. Park, B.H. Kim, Y.-W. Yi, T. Kim, T.K. Kim, I.H. Lee, S.R. Paik, T.
Hyeon, J. Am. Chem. Soc. 128 (2006) 10658–10659.
[
[
21] C.M.R. Rem e´ dios, J.M. Sasaki, Powder Diffr. 23 (2008) S56–S58.
22] H. Shim, P. Dutta, M.S. Seehra, J. Bonevich, Solid State Commun. 145 (2008) 192–
196.
[23] (a) S. Lefondeur, S. Monteverdi, S. Molina, M.M. Bettahar, Y. Fort, E.A. Zhilinskaya,
A. Aboukais, M. Lelaurain, J. Mater. Sci. 36 (2001) 2633–2638;
(b) A.A. Konchits, F.V. Motsnyi, Y.N. Petrov, S.P. Kolesnik, V.S. Yefanov, M.L.
Terranova, E. Tamburri, S. Orlanducci, V. Sessa, M. Rossi, J. Appl. Phys. 100
(
(
2006) 124315/1–124315/7;
c) N. Guskos, M. Maryaniak, J. Typek, P. Podsiadly, U. Narkiewicz, E. Senderek, Z.
Acknowledgements
Roslaniec, J. Non-Cryst. Solids 355 (2009) 1400–1404.
[
24] Y. Moglie, Doctoral Thesis Dissertation, Bah ı´ a Blanca, Argentina, 2009.
This work was generously supported by the Spanish Ministerio
de Educaci o´ n y Ciencia (MEC; grant no. CTQ2007-65218 and
Consolider Ingenio 2010-CSD2007-00006) and the Generalitat
Valenciana (grant no. PROMETEO/2009/039). P. R. thanks the MEC
for a predoctoral grant. We are grateful to Javier Ruiz-Mart ı´ nez for
carrying out the TPO analysis at the Inorganic Chemistry
Department of the University of Alicante.
[25] M.J. Andrews, C.N. Pillai, Indian J. Chem. B 16 (1978) 465–468.
[
26] J. Wettergren, E. Buitrago, P. Ryberg, H. Adolfsoon, Chem. Eur. J. 15 (2009) 5709–
718.
5
[27] (a) C.F. de Graauw, J.A. Peters, H. van Bekkum, J. Huskens, Synthesis (1994) 1007–
1017;
(
(
b) K. Nishide, M. Node, Chirality 14 (2002) 759–767;
c) J.S. Cha, Org. Proc. Res. Dev. 10 (2006) 1032–1053.
[
[
28] J.J. Spivey, G.W. Roberts, B.H. Davies (Eds.), Catalyst Deactivation, Elsevier,
Amsterdam, 2001.
29] (a) A. Shamsi, Appl. Catal. A: Gen. 277 (2004) 23–30;
(
b) Y.-X. Pan, C.-J. Liu, P. Shi, J. Power Sources 176 (2008) 46–53.
References
[
[
[
30] U.K. Singh, S.W. Krska, Y. Sun, Org. Process Res. Dev. 10 (2006) 1153–1156.
31] C. Li, T.C. Brown, Carbon 39 (2000) 725–732.
32] See, for instance: E. Ochoa-Fern a´ ndez, D. Chen, Z. Yu, B. Tøtdal, M. Rønning, A.
Holmen, Catal. Today, 102/103 (2005) 45–49.
[
[
1] W. Tseng, C. Chen, J. Mater. Sci. 41 (2006) 1213–1219.
2] (a) T. Hyeon, Chem. Commun. (2003) 927–934;
(
(
b) S.P. Gubin, Y.A. Koksharov, G.B. Khomutov, G.Y. Yurkov, Russ. Chem. Rev. 74
2005) 489–520.
[
33] A. Tanksale, J.N. Beltramini, J.A. Dumesic, G.Q. Lu, J. Catal. 258 (2008) 366–377.

F. Alonso et al. / Applied Catalysis A: General 378 (2010) 42–51
51
[
[
[
34] (a) Y. Rezgui, M. Guemini, Ing. Eng. Chem. Res. 47 (2008) 4056–4060;
b) A. Benamar, Z. Bechket, Y. Boucheffa, A. Miloudi, C. R. Chim. 12 (2009) 706–
15.
35] (a) J. Li, G. Xiong, Z. Feng, Z. Liu, Q. Xin, C. Li, Micropor. Mesopor. Mater. 39 (2000)
75–280;
b) Y.T. Chua, P.C. Stair, J. Catal. 213 (2003) 39–46.
36] (a) M. Digne, K. Marchand, P. Bourges, Oil Gas Sci. Technol.: Rev. IFP 62 (2007)
1–99;
[37] (a) A.V. Gaikwad, A. Holuigue, M.B. Thathagar, J.E. ten Elshof, G. Rothenberg,
Chem. Eur. J. 13 (2007) 6908–6913;
(
7
(b) L. Dur a´ n Pach o´ n, G. Rothenberg, Appl. Organomet. Chem. 22 (2008) 288–299.
[38] (a) J.A. Widegren, R.G. Finke, J. Mol. Catal. A: Chem. 198 (2003) 317–341;
(b) N.T.S. Phan, M. VanderSluys, C.W. Jones,Adv. Synth. Catal. 348(2006) 609–679;
(c) M. G o´ mez, I. Favier, in: B. Corain, G. Schmid, N. Toshima (Eds.), Metal Nanoclus-
ters in Catalysis and Materials Science: The Issue of Size Control, Elsevier, Amster-
dam, 2008(chapter 31).
2
(
9
(
(
b) G. Katumba, B.W. Mwakikunga, T.R. Mothibinyane, Nanoscale Res. Lett. 3
2008) 421–426.
[39] (a) K.C. Campbell, J.S. Hislop, J. Catal. 13 (1969) 12–19;
(b) J. Dunleavy, Platinum Metals Rev. 50 (2006) 156.